A FICTITIOUS NOISE COMPENSATING TECHNIQUE OF ADAPTIVE KALMAN FILTERING FOR SYSTEMS WITH MODEL ERRORS

In this paper the problem of adaptive Kalman filtering for systems with model errors is discussed.A new technique that the model errors are compensated by introducting fictitious noises is presented by using adaptive Kalman filtering for systems with unknown timevarying noise statistics.Three examples of simulation of different kinds aregiven to show the usefulness of the proposed technique.